开云体育

On Mon, Mar 20, 2023 at 02:26 PM, Roger Need wrote:

50 ohm is not a common value

49.9Ω is a standard E96 1% resistor value. There is an advantage to using two 100Ω resistors in parallel, which is decreased inductance, especially if they are arranged 180° apart to minimize mutual coupling between the resistors.
73, Don N2VGU

On 3/21/23 6:42 AM, Donald S Brant Jr wrote:

On Mon, Mar 20, 2023 at 02:26 PM, Roger Need wrote:

50 ohm is not a common value

49.9Ω is a standard E96 1% resistor value. There is an advantage to using two 100Ω resistors in parallel, which is decreased inductance, especially if they are arranged 180° apart to minimize mutual coupling between the resistors.
73, Don N2VGU

I think you meant 90 degrees apart...

I'm not sure that the reduction in Mutual L wouldn't be countered by the increased trace length. Either way, we're talking nanohenries.

On Tue, Mar 21, 2023 at 09:56 AM, Jim Lux wrote:

I think you meant 90 degrees apart...Either way, we're talking nanohenries.

No, I meant 180. 1nanoHenry at 450MHz is 2.83Ω, which is fairly significant. 0.19Ω at 3.0MHz. It all depends on your frequency and your application.
73, Don N2VGU

Comments relating to recent posts:

1. @Roger do you have a photo of your S21 jig? (By the way, I initially thought your S11 jig was copper sheet with a non-conductive backing board to attach the copper sheet to -- instead you clarified it is copper-clad PCB. Same idea I guess, but on a different scale.)
2. @Roger - how will you make the > < V's you mentioned? I assume the points of the V's won't touch, but will be close to each other. Does the point need to be centered vertically or could one surface just be cut down with the corner of a file to a sort-of V shape?
3. @Donald, I don't understand this 180 degree positioning. Would that be end-to-end? How much spacing (in mm) between them?

On Tue, Mar 21, 2023 at 04:40 PM, Steve Johnson wrote:

1. @Roger do you have a photo of your S21 jig? (By the way, I initially
thought your S11 jig was copper sheet with a non-conductive backing board to
attach the copper sheet to -- instead you clarified it is copper-clad PCB.
Same idea I guess, but on a different scale.)

I do not have a photo of my S21 jig. Basically it is a 50 ohm strip line with SMA connectors on both ends. The stripline has a break in the middle. Pin jacks near the center are used to accept the S11 cal loads and for the S21 through and isolation. It works well into the hundreds of MHz. A better design will use multiple identical boards with soldered on resistors.

2. @Roger - how will you make the > < V's you mentioned? I assume the points
of the V's won't touch, but will be close to each other. Does the point need
to be centered vertically or could one surface just be cut down with the
corner of a file to a sort-of V shape?

I will use a PCB layout. Easy to do and much nicer to work with. The purpose of the V arrangement is to reduce the capacitance between the two sides. Right now it is a little less than 1 pF. Attached is a poor mockup of what the improved board will look like. The layout for the Short, Open and Load is the same on both sides. What is not shown is multiple vias from one side to the other to connect the two sides together. SMD resistors are installed on Load board. Short board will use a couple of 0 ohm resistors or a shorting bridge.

Note that this jig is only useful up to 60 MHz. or so from my previous tests. There is quite an impedance bump due to the BNC to binding posts that has an effect at higher frequencies. I only use this for testing bigger components. For SMD parts I use another pin jack jig which works at much higher frequencies.

3. @Donald, I don't understand this 180 degree positioning. Would that be
end-to-end? How much spacing (in mm) between them?

The position and spacing of the resistors is not critical at the frequencies used by this simple jig. The inductance of the 0805 resistors is only a fraction of a nH and there is already considerable inductance on the copper board connecting the binding posts anyway.

Attached are a couple of plots made with this jig. One is a resistor they measured 557 ohms @ DC and the other is a ceramic cap marked 68 pF.

Roger

On Tue, Mar 21, 2023 at 07:40 PM, Steve Johnson wrote:

@Donald, I don't understand this 180 degree positioning. Would that be
end-to-end? How much spacing (in mm) between them?

Yes, end to end at the end of your transmission line, the free ends going to a ground plane surrounding the end of the line.
Think of a baseball diamond, the transmission line runs from home base to the pitcher's mound; the resistors point from there to first and third base, the grass is your ground.

I build them into the connector; I put them directly across the connector with the resistors going from center conductor to ground. Make a short by soldering a piece of shim stock across the connector, with a hole to solder the center conductor. Use an open connector with the center conductor cut flush for the open circuit.
73, Don N2VGU

One picture would be worth a thousand words.

Thanks,? Doug, VA3DKA

--
This email has been checked for viruses by Avast antivirus software.
www.avast.com

I use this calibration setup:

and for measuring the DUT:



Op wo 22 mrt 2023 om 15:50 schreef Doug <jdkearney@...>:

Thanks for the info. I am only interested in frequencies up to 60 mHz (typical ham radio stuff on HF bands). I’ll have to order those 805 surface mount resistors. I have some old-school metal film leaded 100 ohm resistors— maybe I could use those if I trim the leads very short?

On Wed, Mar 22, 2023 at 11:00 AM, Steve Johnson wrote:

Thanks for the info. I am only interested in frequencies up to 60 mHz
(typical ham radio stuff on HF bands). I’ll have to order those 805 surface
mount resistors. I have some old-school metal film leaded 100 ohm
resistors— maybe I could use those if I trim the leads very short?

The metal film leaded have some internal inductance due to the way they are trimmed.

Roger

Note that all of this depends on how accurate you need impedance
measurements to be.
The metal film resistors will work reasonably for calibration if all you
need is general log-mag S11, SWR, or looking at the general shape and value
of the loss curve / impedance curve for a choke or core. Especially at
frequencies less than 30MHz. If you need accurate impedance measurements,
then the calibration accuracy is more important.
I would order the 805's, but start with the metal film. Then when the 805s
arrive, do a comparison of the difference in measurement you get with the
new calibration. Depending on your specific use case, it may not change
much. Note that in the banana-plug test fixture you plan to use, there are
lots of non-ideal factors at play - so making one factor ideal may not have
as much effect as hoped.
Stan

On Wed, Mar 22, 2023 at 11:06?AM Roger Need via groups.io <sailtamarack=
[email protected]> wrote:

On Wed, Mar 22, 2023 at 11:00 AM, Steve Johnson wrote:

Thanks for the info. I am only interested in frequencies up to 60 mHz
(typical ham radio stuff on HF bands). I’ll have to order those 805

surface

mount resistors. I have some old-school metal film leaded 100 ohm
resistors— maybe I could use those if I trim the leads very short?

The metal film leaded have some internal inductance due to the way they
are trimmed.

Roger

We have no allocation in the milliHertz region (mHz)! That's the land of
earthquakes and the like. If you meant amateur radio HF allocation, that
would be in the MegaHertz region (MHz).

Dave - W?LEV

On Wed, Mar 22, 2023 at 6:00?PM Steve Johnson <cascadianroot@...>
wrote:

Thanks for the info. I am only interested in frequencies up to 60 mHz
(typical ham radio stuff on HF bands). I’ll have to order those 805 surface
mount resistors. I have some old-school metal film leaded 100 ohm
resistors— maybe I could use those if I trim the leads very short?

--
*Dave - W?LEV*


--
Dave - W?LEV

milli...mega...who's counting...

New lottery "Milli Millions"! Think of what you could do with millis!

Don't you mean a mill - a tenth of a cent?

Mike N2MS

OK, OK...

Would that be "Mega Mills" or "Milli Mills"???

On Tue, Mar 21, 2023 at 04:40 PM, Steve Johnson wrote:

Comments relating to recent posts:

1. @Roger do you have a photo of your S21 jig?

Attached is a photo of one I made for doing S21 measurements. It is made using one of the sections of the NanoVNA RF Test Boards that you can find on eBay, Amazon and AliExpress. Works OK for testing things like crystals, RF chokes etc. At higher frequencies the use of the edelay port extension feature will move the S11 reference plane closer to the DUT pins

Roger

Roger, for my purposes, your jig is fantastic. Thamkyou.

73
Mike N7ATC

Finally made my NanoVNA DUT calibration fixture. I cut the gap too wide on the load board so had to "fill in" with a piece of copper wire. The tiny little surface-mounted resistors were a bear to get soldered, but ended up (via luck) getting it done. Lost 2-3 of them before getting lucky.

Very nice setup Steve, thank you for posting!
I may steal your idea,...
But why the 90degree conector and the extra adapters? can't you use just the SMA cable on the BNC adapter? It looks like some unneeded faulty points, I'm guessing it is for some particular situation where you need that corner done.

Cheers,
Pedro On Thursday, May 25, 2023, 03:09:45 AM GMT+1, Steve Johnson <cascadianroot@...> wrote:

Finally made my NanoVNA DUT calibration fixture. I cut the gap too wide on the load board so had to "fill in" with a piece of copper wire.? The tiny little surface-mounted resistors were a bear to get soldered, but ended up (via luck) getting it done. Lost 2-3 of them before getting lucky.

On Sun, Mar 12, 2023 at 08:25 PM, Don Rolph wrote:

Fahnstock clips no longer exist,

They exist on Amazon -

(lots of them!)

And yes, I'm surprised!
--
Doug, K8RFT